doinkboy
@doinkboy.bsky.social
Social Science Data Analysis (especially Dyadic, Mediation, SEM; davidakenny.net) and My Twisted Take on Political & Cultural Topics; Reposts Are Not Necessarily Endorsements
Reposted by doinkboy
“Meta shut down internal research into the mental health effects of Facebook after finding causal evidence that its products harmed users’ mental health”
It's worth keeping in mind with both of these approaches that content-mediated harm may not be picked up by broad patterns of use. This is what Meta seems to have sorted out internally as the key mechanism, and then promptly buried.
www.reuters.com/sustainabili...
www.reuters.com/sustainabili...
Meta buried 'causal' evidence of social media harm, US court filings allege
Rather than publishing the findings or pursuing additional research, Meta called off further work.
www.reuters.com
January 7, 2026 at 5:11 PM
“Meta shut down internal research into the mental health effects of Facebook after finding causal evidence that its products harmed users’ mental health”
Reposted by doinkboy
Pleased to share for those using #rstats for analyzing #dyadic data, that we released a major update of dySEM on CRAN (now ver. 1.4.1) *just* before the winter break.
New functionality for scripting/outputting dyadic SEMs, along with (much) more test coverage! 🧵
cran.r-project.org/web/packages...
New functionality for scripting/outputting dyadic SEMs, along with (much) more test coverage! 🧵
cran.r-project.org/web/packages...
January 7, 2026 at 4:49 PM
Pleased to share for those using #rstats for analyzing #dyadic data, that we released a major update of dySEM on CRAN (now ver. 1.4.1) *just* before the winter break.
New functionality for scripting/outputting dyadic SEMs, along with (much) more test coverage! 🧵
cran.r-project.org/web/packages...
New functionality for scripting/outputting dyadic SEMs, along with (much) more test coverage! 🧵
cran.r-project.org/web/packages...
Reposted by doinkboy
As a statistical educator, it had not occurred to me that I need to caution students against regressing a variable on a function of itself. My naivete is unbounded.
The Peri & Sparber paper (linked below) looks really good! It has synthetic data analyses and everything.
The Peri & Sparber paper (linked below) looks really good! It has synthetic data analyses and everything.
Third, it's true: some immigration researchers' choices do bias their results.
A key example is Borjas (2006).
Peri & Sparber (2011) show that Borjas chose a regression specification that generated spurious negative correlation between immigration & native employment.
doi.org/10.1016/j.ju...
A key example is Borjas (2006).
Peri & Sparber (2011) show that Borjas chose a regression specification that generated spurious negative correlation between immigration & native employment.
doi.org/10.1016/j.ju...
January 8, 2026 at 9:16 AM
As a statistical educator, it had not occurred to me that I need to caution students against regressing a variable on a function of itself. My naivete is unbounded.
The Peri & Sparber paper (linked below) looks really good! It has synthetic data analyses and everything.
The Peri & Sparber paper (linked below) looks really good! It has synthetic data analyses and everything.
Reposted by doinkboy
Reposted by doinkboy
The 1.3.0 version of my {brms} + {tidyverse} translation of Kruschke's "Doing Bayesian data analysis" is up!
solomon.quarto.pub/dbda2
#rstats
1/8
solomon.quarto.pub/dbda2
#rstats
1/8
Doing Bayesian Data Analysis in brms and the tidyverse
solomon.quarto.pub
January 2, 2026 at 5:00 PM
The 1.3.0 version of my {brms} + {tidyverse} translation of Kruschke's "Doing Bayesian data analysis" is up!
solomon.quarto.pub/dbda2
#rstats
1/8
solomon.quarto.pub/dbda2
#rstats
1/8
Reposted by doinkboy
And we're live, Lecture A1 is online. Introduction to Bayesian workflow, generative models, estimands, estimators, estimates, error checking, beginnings of probability theory and Bayesian updating. www.youtube.com/watch?v=ztbY...
January 6, 2026 at 11:02 AM
And we're live, Lecture A1 is online. Introduction to Bayesian workflow, generative models, estimands, estimators, estimates, error checking, beginnings of probability theory and Bayesian updating. www.youtube.com/watch?v=ztbY...
Reposted by doinkboy
Not wrong. To use a bit of understatement, stackoverflow was not always the most friendly place
January 5, 2026 at 2:25 PM
Not wrong. To use a bit of understatement, stackoverflow was not always the most friendly place
Reposted by doinkboy
The Problem:
We're measuring wrong. We treat interventions as binary switches: worked or didn't.
But causation isn't binary—it's a PROCESS.
Effects evolve through phases:
→ Strong direct effects early
→ Indirect pathways activate later
→ Hidden mechanisms undermine progress
We're measuring wrong. We treat interventions as binary switches: worked or didn't.
But causation isn't binary—it's a PROCESS.
Effects evolve through phases:
→ Strong direct effects early
→ Indirect pathways activate later
→ Hidden mechanisms undermine progress
January 4, 2026 at 11:00 PM
The Problem:
We're measuring wrong. We treat interventions as binary switches: worked or didn't.
But causation isn't binary—it's a PROCESS.
Effects evolve through phases:
→ Strong direct effects early
→ Indirect pathways activate later
→ Hidden mechanisms undermine progress
We're measuring wrong. We treat interventions as binary switches: worked or didn't.
But causation isn't binary—it's a PROCESS.
Effects evolve through phases:
→ Strong direct effects early
→ Indirect pathways activate later
→ Hidden mechanisms undermine progress
Reposted by doinkboy
Imagine how much good those funds could have done to advance research integrity instead.
January 4, 2026 at 1:00 PM
Imagine how much good those funds could have done to advance research integrity instead.
Venezuela nationalized its oil on 1976. Saudi Arabia nationalized its in 1980. Is the US going to reseize Saudi oil now? #Venezuela #Trump
January 3, 2026 at 5:38 PM
Venezuela nationalized its oil on 1976. Saudi Arabia nationalized its in 1980. Is the US going to reseize Saudi oil now? #Venezuela #Trump
Reposted by doinkboy
So if you’re Venezuelan and in the U.S., we’re kidnapping you and flying you back to Venezuela.
And if you’re Venezuelan and in Venezuela, we’re kidnapping you and flying you to the U.S.
And if you’re Venezuelan and in Venezuela, we’re kidnapping you and flying you to the U.S.
January 3, 2026 at 3:11 PM
So if you’re Venezuelan and in the U.S., we’re kidnapping you and flying you back to Venezuela.
And if you’re Venezuelan and in Venezuela, we’re kidnapping you and flying you to the U.S.
And if you’re Venezuelan and in Venezuela, we’re kidnapping you and flying you to the U.S.
Reposted by doinkboy
Marco Rubio is reportedly saying Maduro will stand trial in US courts.
Which means it’s now the US administration’s position that US courts can hold foreign presidents, but not the US president, accountable for crimes.
Which means it’s now the US administration’s position that US courts can hold foreign presidents, but not the US president, accountable for crimes.
January 3, 2026 at 10:52 AM
Marco Rubio is reportedly saying Maduro will stand trial in US courts.
Which means it’s now the US administration’s position that US courts can hold foreign presidents, but not the US president, accountable for crimes.
Which means it’s now the US administration’s position that US courts can hold foreign presidents, but not the US president, accountable for crimes.
Reposted by doinkboy
TIL the Nazis tried to establish their own version of Silent Night (see alt text for translation).
de.wikipedia.org/wiki/Stille_...
de.wikipedia.org/wiki/Stille_...
January 1, 2026 at 11:23 AM
TIL the Nazis tried to establish their own version of Silent Night (see alt text for translation).
de.wikipedia.org/wiki/Stille_...
de.wikipedia.org/wiki/Stille_...
Reposted by doinkboy
To be clear, the paper doesn't cover measurement error as discussed in the blog post, but I am sharing it anyway
December 30, 2025 at 12:50 PM
To be clear, the paper doesn't cover measurement error as discussed in the blog post, but I am sharing it anyway
Reposted by doinkboy
Looking at the Port Huron Statement, 63 years later
statmodeling.stat.columbia.edu/2025/12/28/l...
statmodeling.stat.columbia.edu/2025/12/28/l...
Looking at the Port Huron Statement, 63 years later | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
December 28, 2025 at 2:16 PM
Looking at the Port Huron Statement, 63 years later
statmodeling.stat.columbia.edu/2025/12/28/l...
statmodeling.stat.columbia.edu/2025/12/28/l...
Reposted by doinkboy
This: In the same week that Vance expressed concern that France and the United Kingdom had nuclear weapons, he said Denmark was not a very good ally to the US. It is shameful.
bsky.app/profile/thom...
bsky.app/profile/thom...
The former ambassador to Denmark for the United States, Rufus Gifford, posted this video on his Facebook account:
Ht: @hpsc24.bsky.social
Ht: @hpsc24.bsky.social
December 27, 2025 at 4:33 PM
This: In the same week that Vance expressed concern that France and the United Kingdom had nuclear weapons, he said Denmark was not a very good ally to the US. It is shameful.
bsky.app/profile/thom...
bsky.app/profile/thom...
Reposted by doinkboy
And in case you think I'm being insensitive, remember that in the wake of school shootings, Kirk did say that they were worth it, to preserve the freedom [sic] guaranteed by the 2nd Amendment
December 25, 2025 at 8:25 PM
And in case you think I'm being insensitive, remember that in the wake of school shootings, Kirk did say that they were worth it, to preserve the freedom [sic] guaranteed by the 2nd Amendment
In the Epstein files, there is a Jan. 7, 2020 email from a prosecutor (name redacted) to someone (possibly Steven Miller, but name redacted) that Trump flew on Epstein's plane several times. Trump had previously said he never had.
December 25, 2025 at 4:16 AM
In the Epstein files, there is a Jan. 7, 2020 email from a prosecutor (name redacted) to someone (possibly Steven Miller, but name redacted) that Trump flew on Epstein's plane several times. Trump had previously said he never had.
Reposted by doinkboy
Survey Statistics: is a mismeasured X better than none at all ?
statmodeling.stat.columbia.edu/2025/12/23/s...
statmodeling.stat.columbia.edu/2025/12/23/s...
Survey Statistics: is a mismeasured X better than none at all ? | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
December 24, 2025 at 10:47 AM
Survey Statistics: is a mismeasured X better than none at all ?
statmodeling.stat.columbia.edu/2025/12/23/s...
statmodeling.stat.columbia.edu/2025/12/23/s...
Reposted by doinkboy
The problems with popular internet heuristics such as “Hanlon’s razor,” “steelmanning,” and “Godwin’s law,” all of which kind of fall apart in the presence of actual malice, actual bad ideas, and actual Nazis.
statmodeling.stat.columbia.edu/2025/12/23/t...
statmodeling.stat.columbia.edu/2025/12/23/t...
The problems with popular internet heuristics such as “Hanlon’s razor,” “steelmanning,” and “Godwin’s law,” all of which kind of fall apart in the presence of actual malice, actual bad ideas, and act...
statmodeling.stat.columbia.edu
December 23, 2025 at 3:07 PM
The problems with popular internet heuristics such as “Hanlon’s razor,” “steelmanning,” and “Godwin’s law,” all of which kind of fall apart in the presence of actual malice, actual bad ideas, and actual Nazis.
statmodeling.stat.columbia.edu/2025/12/23/t...
statmodeling.stat.columbia.edu/2025/12/23/t...
Reposted by doinkboy
youtu.be/fB2TK58nV5Q?.... DOJ leaked evidence that they previously claimed not to exist #Bondi #DOJ #EpsteinFiles
New Epstein files accidentally leaked and it's bad for Trump...
YouTube video by Keith Edwards
youtu.be
December 23, 2025 at 5:54 AM
youtu.be/fB2TK58nV5Q?.... DOJ leaked evidence that they previously claimed not to exist #Bondi #DOJ #EpsteinFiles
When will Trump ask to rename Washington DC as Trump DC?
December 19, 2025 at 3:50 PM
When will Trump ask to rename Washington DC as Trump DC?
Reposted by doinkboy
I've used Lavaan almost every (work)day for 15 years, but this open source labour of love has never received proper institutional support. I'm delighted to be a small part of an 1.5M OpenScienceNL award, led by Jorgensen, to completely revamp and futureproof Lavaan www.openscience.nl/en/news/45-p...
December 18, 2025 at 7:44 AM
I've used Lavaan almost every (work)day for 15 years, but this open source labour of love has never received proper institutional support. I'm delighted to be a small part of an 1.5M OpenScienceNL award, led by Jorgensen, to completely revamp and futureproof Lavaan www.openscience.nl/en/news/45-p...
Reposted by doinkboy
Just gave my last talk of the year!
2025 was quite packed, I gave talks about:
- the age-period-cohort problem
- making rigorous causal inference more mainstream
- mediation analysis
- marginaleffects
- causal graphs (x10)
If you're curious, check out my slides here: juliarohrer.com/resources/
2025 was quite packed, I gave talks about:
- the age-period-cohort problem
- making rigorous causal inference more mainstream
- mediation analysis
- marginaleffects
- causal graphs (x10)
If you're curious, check out my slides here: juliarohrer.com/resources/
Resources
Here you can find a collection of things that may be helpful, including slide decks, a curated list of introductory papers and blog posts, as well as some infographics I have generated to explain v…
juliarohrer.com
December 18, 2025 at 12:38 PM
Just gave my last talk of the year!
2025 was quite packed, I gave talks about:
- the age-period-cohort problem
- making rigorous causal inference more mainstream
- mediation analysis
- marginaleffects
- causal graphs (x10)
If you're curious, check out my slides here: juliarohrer.com/resources/
2025 was quite packed, I gave talks about:
- the age-period-cohort problem
- making rigorous causal inference more mainstream
- mediation analysis
- marginaleffects
- causal graphs (x10)
If you're curious, check out my slides here: juliarohrer.com/resources/
Reposted by doinkboy
Oh Canada!
Employment growth since April ("Liberation Day"):
United States: +119,000*
Canada: +166,200
* The U.S. number is expected to be revised down.
United States: +119,000*
Canada: +166,200
* The U.S. number is expected to be revised down.
December 18, 2025 at 1:26 AM
Oh Canada!